Dynamic Image Segmentation based Full-waveform Inversion of Subsurface Penetrating Radar Data

Full-waveform inversion (FWI) has become the centerpiece of velocity model building (VMB) in seismic data processing in recent years. It has proven capable of significantly improving the velocity model and, thus, the migration image for different acquisition types and geologic settings, including complex environments such as salt. With the advent of FWI imaging, the scope of FWI applications has extended further from VMB into the imaging landscape. However, current FWI applications in the industry prevalently employ the acoustic approximation. One common problem of acoustic FWI (A-FWI) is the apparent salt halos at the salt-sediment interface in the resulting FWI velocity and FWI image, the presence of which hinders direct interpretation and imaging focusing around salt bodies. With synthetic and field data examples, we demonstrate that this salt halo is caused mainly by the large mismatch between the elastic recorded data and the acoustic modeled data, particularly at middle to long offsets. To overcome limitations imposed by acoustic assumptions, we developed an elastic FWI (E-FWI) algorithm that combines an elastic modeling engine with the time-lag cost function, which we call elastic time-lag FWI (E-TLFWI). With a more accurate modeling engine, E-TLFWI significantly reduces the salt halo observed in its acoustic counterpart. However, we also observe that the images migrated using the A-FWI and E-FWI velocity models remain similar overall, with some slight improvements around and beneath salt boundaries, particularly near steep salt flanks, as a result of the reduced salt halo. By contrast, FWI images derived from E-TLFWI show considerable benefits over those from acoustic time-lag FWI, such as improved event focusing, better structural continuity, and higher signal-to-noise ratio. The sharpened salt boundaries and enhanced quality of the FWI images reveal the significant value of E-FWI and provide the justification for its greatly increased cost.

The image of biometric properties in humans is used in many fields today. Regardless of these features, it is necessary to first translate it into data that the computer understands. In this study, automatic and dynamic image segmentation was performed by using 300x300 fingerprint images. A fingerprint database with a total of 80 images and 10 different classes was used. The features of the images were subtracted from the sub-segments obtained from these images by the feature extraction algorithm that was originally developed. The 300x300 images were divided into 25x25 sub-images and the feature vector was obtained. 144x80 inputs obtained after image segmentation were kept in areas in separate tables. The developed segmentation and feature extraction algorithm can be applied to any image of equal size.

With the increasing demand in complexity for subsurface models in environments such as subsalt, sub-basalt and pre-salt, full-waveform inversion (FWI) is quickly becoming one of the model-building methods of choice. While in principle capable of handling all of the nonlinearity in the data, in practice nonlinear gradient-based FWI is limited due to its notorious sensitivity to the choice of starting models. To help addressing model convergence issues in FWI, in this paper we analyze the role of nonlinearity in the so-called sensitivity kernels, which are the centerpiece of gradient-based FWI algorithms. Using a scattering-based approach and assuming acousticonly data, we start by reparameterizing the subsurface model in terms of smooth and singular components for both compressibility and density. This leads to a decomposition of the data into a reference field that is sensitive only to the smooth model, and a scattered field sensitive to both smooth and sharp model components. Focussing on the model backprojections from the scattered data only, we provide expressions for the Fr´echet-derivative sensitivity kernels of all four model parameters. Our results provide for the decomposition of current FWI kernels into no less than nine different sub-kernels which have explicitly different levels of nonlinearity with respect to both data and model parameters. This capability to discern levels of nonlinearity within FWI kernels is key to understanding model convergence in gradient-based, iterative FWI. We illustrate this by analyzing some of the sub-kernel terms in detail. The scattering-based FWI kernel decomposition we provide could have broad potential applications, such devising multiscale FWI algorithms, and improving velocity model building in the image domain using extended image gathers.

Electron beam computed tomography is unparalleled in its ability to consistently produce high quality dynamic images of the human heart. Its use in quantification of left ventricular dynamics is well established in both clinical and research applications. However, the image analysis tools supplied with the scanners offer a limited number of analysis options. They are based on embedded computer systems which have not been significantly upgraded since the scanner was introduced over a decade ago in spite of the explosive improvements in available computer power which have occured during this period. To address these shortcomings, a workstation-based ventricular analysis system has been developed at our institution. This system, which has been in use for over five years, is based on current workstation technology and therefore has benefited from the periodic upgrades in processor performance available to these systems. The dynamic image segmentation component of this system is an interactively supervised, semi-automatic surface identification and tracking system. It characterizes the endocardial and epicardial surfaces of the left ventricle as two concentric 4D hyper-space polyhedrons. Each of these polyhedrons have nearly ten thousand vertices which are deposited into a relational database. The right ventricle is also processed in a similar manner. This database is queried by other custom components which extract ventricular function parameters such as regional ejection fraction and wall stress. The interactive tool which supervises dynamic image segmentation has been enhanced with a temporal domain display. The operator interactively chooses the spatial location of the endpoints of a line segment while the corresponding space/time image is displayed. These images, with content resembling M-Mode echocardiography, benefit form electron beam computed tomography's high spatial and contrast resolution. The segmented surfaces are displayed along with the imagery. These displays give the operator valuable feedback pertaining to the contiguity of the extracted surfaces. As with M-Mode echocardiography, the velocity of moving structures can be easily visualized and measured. However, many views inaccessible to standard transthoracic echocardiography are easily generated. These features have augmented the interpretability of cine electron beam computed tomography and have prompted the recent cloning of this system into an 'omni-directional M-Mode display' system for use in digital post-processing of echocardiographic parasternal short axis tomograms. This enhances the functional assessment in orthogonal views of the left ventricle, accounting for shape changes particularly in the asymmetric post-infarction ventricle. Conclusions: A new tool has been developed for analysis and visualization of cine electron beam computed tomography. It has been found to be very useful in verifying the consistency of myocardial surface definition with a semi-automated segmentation tool. By drawing on M-Mode echocardiography experience, electron beam tomography's interpretability has been enhanced. Use of this feature, in conjunction with the existing image processing tools, will enhance the presentations of data on regional systolic and diastolic functions to clinicians in a format that is familiar to most cardiologists. Additionally, this tool reinforces the advantages of electron beam tomography as a single imaging modality for the assessment of left and right ventricular size, shape, and regional functions.

Full waveform inversion (FWI) is an optimization procedure which minimizes the waveform difference between the observed and calculated seismic data. The method, through an iterative forward modeling and inversion based on the acoustic or elastic wave equation, has recently gained wide interest with specific application in exploration and producing. FWI is wellknown for its ability to obtain high resolution and accurate subsurface models in areas with complex geological structures related to the presence of salt bodies encountered in deep water exploration. However, conventional FWI algorithms require seismic data that contains low frequencies (<5Hz). Lack of low frequencies results in unreliable velocity models for seismic imaging. In this paper, we aim to relax the low-frequency requirement of FWI and propose a technique which employs principles from sparse deconvolution in order to recover the missing low frequencies. Through a series of synthetic examples, it is shown that our proposed methodology is able to preserve and utilize the traveltime information of the original data; hence, it allows convergence of the FWI. The major advantage of this method is that it is fully automatic, involving no residual moveout or traveltime difference analysis.

SUMMARY The inversion of complete seismic waveforms offers new perspectives to better constrain the elastic properties of Earth’s interior. However, models of density and seismic velocities obtained from full waveform inversions are generally characterized by very different and uneven spatial resolutions. Because the 3-D structure of the Earth represents small deviations from average reference Earth models, the absolute values of density, VP and VS in the Earth are strongly correlated. Here, we exploit this strong correlation between model parameters as a priori information introduced into a new full waveform inversion algorithm, by considering a non-diagonal 3-D model covariance matrix in which the spatial correlations of elastic properties are described with an exponential covariance function. The inverse of such a model covariance matrix is easy to compute, and we thus have all the ingredients to construct a consistent Bayesian full waveform inversion scheme. We show that taking into account the correlations between density and seismic velocities can lead to dramatic improvements on the reconstructed models of density, seismic velocities and VP/VS ratio. This new imaging approach opens new perspectives for refining tomographic images of density and seismic velocities in the lithosphere and upper mantle on a regional scale by full waveform inversion of teleseismic body waves.

Summary Ultra-long offset stationary-receiver acquisitions implemented with multi-component Ocean Bottom Nodes (OBN) are emerging as a suitable technology for deep offshore subsalt imaging by full waveform inversion (FWI). Ultra-long offsets allow to undershoot the deepest targeted structures with diving waves. In this framework, first-arrival traveltime + slope tomography (FASTT) provides a simple and efficient tool to build kinematically-accurate initial velocity model for full waveform inversion (FWI). In particular, the incorporation of slopes in the first-arrival tomography in addition to traveltimes removes artifacts in poorly illuminated areas and better reconstruct structural dips. Either the FASTT or the FWI velocity models can be used as reliable background models for prestack depth migration of co-incident towed-streamer data. However, the high resolution of the FWI velocity model facilitates the integrated interpretation of the reflectivity mapped by migration and the intermediate-scale velocity variations reconstructed by waveform inversion. We illustrate the workflow combining first-arrival tomography + FWI of ultra-long offset OBN data and migration of co-incident towed-streamer data with a challenging deep crustal case study in the eastern Nankai subduction zone.

Quadtree representation presents many advantages in image analysis because of its better manipulation at different scales. In computer vision one frequently faces the problem of multi-scale analysis. We use a syntactic representation of quadtree to analyse images in the Haar base with pyramidal algorithms in linear time. We present applications in dynamic filtering, in compression and image segmentation.

Seismic acquisitions in urban areas represent a challenge due to the noisy environment during the recording of the data. This often prevents obtaining high-resolution images of the subsurface in the area of interest. One way to achieve better results, despite the challenging circumstances, is full-waveform inversion (FWI). This method iteratively updates a starting model until the simulated data and the observed field data matches to a satisfactory level. This eventually yields the desired subsurface model. However, FWI comes with some major limitations. Because each iteration requires a calculation of the wave propagation several times for each shot, it is highly computationally expensive. Furthermore, due to storage limitations, especially if high performance computing (HPC) clusters are not available, the model size is restricted as well. This work aims to find a solution for these limitations in an elastic 2D case by implementing the random-shot workflow (RSW). In a conventional FWI workflow which is called full-shot workflow (FSW) in this work, each iteration uses all shots of the recorded data before updating the whole model at once which leads to the high computational and storage costs. RSW randomly picks only one shot, applies FWI within the area of the shot and its corresponding receivers and then updates the model only within this subset. This should, in my expectation, split the workload into several smaller subsets without sacrificing the effectiveness of FWI. After the implementation, I compare RSW to FSW in a synthetic benchmark where I verify the performance. Afterwards, the workflows are compared to each other by using a fraction of a field data set that was acquired in Salt Lake City. These data represents a large urban seismic land data set with a low signal-to-noise ratio. They also have suboptimal conditions such as the end-on spread geometry due to the use of a land streamer and occasionally bad coupling of the geophones. This turns out to be a big challenge for the FWI algorithm. The starting model used for the field data inversion is based on the multichannel analysis of surface waves which deals with the inversion of surface-wave dispersion curves. It is a cost-efficient way to obtain an initial model for FWI.With a subset of this model and the data, I investigate the influence of traces near the sources which suffer heavily from the suboptimal acquisition conditions, and if more iterations can improve the results. Removing such traces improve the major part of the results, while more iterations are not beneficial with this particular data set. At the end I show a final comparison of both workflows by applying them to the complete Salt Lake City field data set. An application of FWI on such a large-scale land streamer acquisition with such a large number of shots has, to the best of my knowledge, not been done yet, most likely because of the large computational costs, even in 2D. The comparisons and investigations show that while RSW has a major computational advantage in storage usage and per-CPU performance, it is too sensitive to bad traces and shots which significantly degrade the final model. Therefore, it is unsuitable for field data with a low signal-to-noise ratio or that have not been preprocessed well. Furthermore, RSW cannot utilise the whole power of HPC clusters while FSW is able to do so by optimal shot parallelisation. Thus, FSW could potentially outperform RSW by raw computing power if such excessive power is available.

Full waveform inversion using a decomposed single frequency component from a spectrogram

Ultrasound computed tomography based on full waveform inversion with source directivity calibration

We deal with the conception of a perception system that's goal is to assist a mobile robot teleoperator by providing pertinent information about eventual obstacles appearing in the robot work space. This range image based perception system is to be embedded on a vehicle able to move at speeds up to 40 km/h in an outdoor environment. A method taking speed constraints into account is proposed. In the first step of this method, a segmentation algorithm is applied to the first range image scanned by the motionless robot to determine areas of interest. From these areas, distinctive attributes are computed and recorded as symbolic representations of each obstacle region. In the second and following steps, obstacles are localized in images scanned during robot motion. The difference between actual object position in the range image and its predicted value is used by an extended Kalman filter to correct the estimated robot configuration. A dynamic image segmentation using emergency and security criteria is carried out and new obstacles can now be detected from the range image and expressed in the robot coordinates system.

CO2 Characterization at the Sleipner Field with Full Waveform Inversion: Application to Synthetic and Real Data

SUMMARY Full waveform inversion (FWI) is a high-precision subsurface imaging technique that inverts subsurface parameter models by minimizing the discrepancy between observed and synthetic seismic data. However, complicated wave propagation mechanisms, non-convexity of the loss function and limited seismic acquisition system necessitate the incorporation of sufficient prior and physical constraints to alleviate the ill-posedness and cycle-skipping. Although the implicit FWI (IFWI) can encode implicit spectral bias (i.e. inverting model parameters from low frequencies to high frequencies) to reduce the dependence on an accurate initial model, its limited high-frequency inversion capability results in thousands of iterations for the final results. In this paper, we indicate that the frequency hyperparameters of the sine activation function in IFWI modulate the spectral bias, making a trade-off between inversion accuracy and stability, that is, lower frequencies yield robust FWI but lower accuracy, while higher frequencies achieve higher accuracy on the premise of an accurate initial model. To improve both the stability and accuracy of IFWI, we propose a novel implicit FWI method with an adaptive Fourier reparametrization strategy (termed FR-IFWI), which explicitly encodes multifrequency information by reparametrizing the network weights using a learnable coefficient matrix and fixed Fourier frequency bases. The role of learnable matrices in neural networks can evolve from determining frequencies in IFWI to actively selecting frequencies from fixed frequency bases through FR-IFWI, which alleviates the dependence on activation function frequency and obtains more robust and accurate inversion results. Extensive numerical experiments on the modified Marmousi, 2-D SEG/EAGE Salt and Overthrust models confirm that FR-IFWI successfully achieves superior inversion efficiency and accuracy compared with conventional FWI and IFWI methods.

After many years of processing seismic streamer data acquired over the Shenzi field in deep-water Gulf of Mexico (GOM) with the best available technology, the workflows have remained highly interpretive, far from robust, often time-consuming, and ultimately proved inadequate for resolving the complexity of the salt and subsalt velocity models. Recent advances in full-waveform inversion (FWI) algorithms combined with a newly acquired, ocean bottom node (OBN) data set have ushered in a fundamental shift in velocity model building and imaging at Shenzi as we move away from mostly interpretation-driven to more data-driven processes. While the velocity model obtained from Time-lag FWI (TLFWI) using the legacy rich-azimuth (RAZ) streamer data provided significant imaging uplift over the best previous processing, an additional step-change improvement in image quality was achieved by running TLFWI with the OBN data. The OBN TLFWI model combined with the long offsets of the OBN data allowed steeply dipping events along the salt feeder to be seen for the first time on the reverse time migration (RTM) image. However, despite these large improvements, the image still suffers in areas of low illumination. Least-squares RTM (LSRTM) showed some good improvement over the RTM but proved less effective where the starting image was too poor. By running TLFWI to a maximum frequency comparable to the RTM, we were able to generate an FWI Image, an estimation of the reflectivity obtained directly from the FWI velocity model, with good S/N, more balanced amplitudes, and improved steep-dip and fault imaging, making it a direct rival to the RTM and LSRTM images.